rocketride-server
ActiveDescription
RocketRide Server is a high-throughput agent runtime designed for production-grade, low-latency LLM tool execution and orchestration.
Key Features
- High-performance AI pipeline engine with a C++ core and 50+ Python-extensible nodes
- Built for RAG and agents with parsing, vectorization, retrieval, and execution nodes
- Composable workflows where nodes are stitched as directed acyclic graphs
- Production observability with built-in metrics, logs, and tracing integration
- Multi-language SDK with Python-first and additional languages supported
Use Cases
💡 Build enterprise RAG systems that handle large-scale document ingestion and retrieval
💡 Compose multiple AI capabilities into observable agent pipelines
💡 Provide a unified pipeline abstraction for multi-model inference
💡 Schedule hybrid RAG + agent workflows under low-latency requirements
Categories
Quick Start
# Install RocketRide
pip install rocketride
# Define a RAG + agent workflow
from rocketride import Pipeline, nodes
pipe = Pipeline()
pipe.add(nodes.DocumentLoader(source="docs/"))
pipe.add(nodes.Embedder(model="text-embedding-3-small"))
pipe.add(nodes.VectorStore(name="kb"))
pipe.add(nodes.AgentNode(llm="claude-sonnet-4"))
# Execute
result = pipe.run(query="Summarize the product FAQ")